Effect of delayed acquisition times on gadolinium-enhanced magnetic resonance imaging of the presumably normal canine brain.

A delay in imaging following intravenous contrast medium administration has been recommended to reduce misdiagnoses. However, the normal variation of contrast enhancement in dogs following a delay has not been characterized. Contrast-enhanced MR imaging of 22 dogs was assessed, in terms of identification of normal anatomic structures, to investigate the variation associated with 10-min delay between contrast medium administration and imaging. All dogs had a normal brain MR imaging study and unremarkable cerebrospinal fluid. Specific regions of interest were assessed both objectively, using computer software, and subjectively using three observers. Mean contrast enhancement > 10% was seen in the pituitary gland, choroid plexus, meninges, temporal muscle, trigeminal nerve, and the trigeminal nerve root. Structures with an active blood-brain barrier had minimal contrast enhancement (< 6%). Enhancing structures had significantly more contrast enhancement at t = 1 min vs. t = 10 min, except in temporal muscle, the trigeminal nerve and the trigeminal nerve root. Interobserver agreement was moderate to good in favor of the initial postcontrast T1-weighted (T1w) sequence. The observers found either no difference or poor agreement in identification of the nonvascular structures. Intraobserver agreement was very good with all vascular structures and most nonvascular structures. A degree of meningeal enhancement was a consistent finding. The initial acquisition had higher enhancement characteristics and observer agreement for some structures; however, contrast-to-noise was comparable in the delayed phase or not significantly different. We provide baseline references and suggest that the initial T1w postcontrast sequence is preferable but not essential should a delayed postcontrast T1w sequence be performed.